"The question, though, is whether Canonical should even be putting much time into desktop builds of Ubuntu — the desktop PC is undoubtedly on its way out..."

Because for the most part all they are is a desktop OS. There were still over 300 million desktop PCs sold last year, and even though it's in decline it is still a huge market. Desktop PCs (including laptops) will never go away completely, and there are still many years ahead where their work will have customers.

Agreed. I'm so tired of people that obviously have very little to contribute to the tech sphere or society writing these incipid articles. I would like the author to write all of his articles from now on, on his cellphone. I would also like for the author to do all photo editing, graphics, game programming and business documents on his cellphone, as well as compose all of his music on his cellphone. Not a tablet.

I think you misunderstood the author, the idea is to have all in one cell phone, and its really great to see Ubuntu taking the lead, all you would need is a big size display a Bluetooth keyboard and mouse to convert your phone into a desktop, so the emphasize would be on phone to make it convertible not the desktop.

That still requires a desktop OS, just one that runs on a mobile device. It will possibly have to shift between mobile/desktop/10ft interfaces, but I'm yet to see a single interface that translates across these (and other) mediums.

You still need work stations. Tablets have come a long way, but for high performance computing, we'll always need work stations. One trend which seems to be occurring is the idea of multiple work stations remotely being connected to a cluster of computers, similar to how people programmed back in the '70s (think UNIX). Either way, while for the average user the desktop is going the way of the dinosaur (in favor of tablets), the majority of professional or hobbyists involved in any form of creative digital media are going to require desktops to accomplish their work. Yes, you can be a web dev with a tablet and do fine. Programming something in OpenGL for x86 CPUs (AFAIK most tablets are ARM-based), however, requires power...and a good GPU, if you want access to a lot of features which will help you write your game.

I can think UNIX - but let's face it, how many people used clusters back in the 70ies, compared to now?

Also, I don't think individual desktop SHOULD ever be lost and that this direction SHOULD indeed be taken by the wide user community - if we end up having only front end stuffs while all the actual work is done on clusters owned by someone else (the service provider), this "someone else" will have us by the unmentionables - whether we're talking data privacy, code privacy, or good ol' service billing.

that's great! :) I'm looking forward to rendering my Rhino creations on a cell phone in 2025! Until then however...........................

The market isn't declining at all; it's the _growth rate_ that is declining. That means that the market keeps growing bigger every day. I have a six year old laptop that I use every day. However, the screen is broken and that is the reason I'm looking for a new one. Otherwise, I could've used it for several years still. The fact that people aren't buying as many new PC as they used to, doesn't mean they don't use PCs.

So long as there are developers, there will be desktop operating systems.

I too seriously disbelieve the "fall" of the desktop/laptop computer in near future, unless some SERIOUSLY new technology is developed (possibly something with holographic display and touch interface). Tablets and phones powered by android or iWhatever are all nice and well - UNTIL you try and get some serious shit DONE. Whether you're a web or graphic designer, a 3D graphics professional, an engineer, a serious software developer, a movie maker or a music professional, I DARE you for a 10-12 hour shift on a 7" screen with 1/3 of it taken by a keyboard... good luck :)

Sorry, Bon Dobbs... The desktop isn't going anywhere. It's the only way that anything actually gets done in this world (work). There is no work to be done with a smartphone with a keyboard... only games and that is rather juvinile.

Trying to do Desktop Work on Smart phones, and a person will loose their eye sight in a few years. Programmers, and people that actually work need desktops, unless they would rather send txt messages.

"the desktop PC is undoubtedly on its way out"

Wut.

" the desktop PC is undoubtedly on its way out"

Not any time in the foreseeable future. Please stop trying to push this idea, it simply isn't true.

The desktop PC *market* is definitely shrinking, so it might be fair to predict that the desktop PC *market* is on its way out. But saying that that desktop PC is *undoubtedly* on its way out is wrong.

Correct it is being augmented, mainly at the consumer level only - where margins are super thin. Business still requires PC's and laptops. A tablet does not remove the need for a desktop or laptop. Even then, most business class tablets could just be considered ultra books.

The perception that the PC market is on its way out is partially due to the fact that the once only PC operating system is no longer relevant and on its way out.

Now that Linux is making inroads in just about every technology on the planet (And elsewhere), there’s going to be a whole new paradigm in computing coming up and PCs are part of that...

Welcome to the new computing world :)

It seems that's where it's been headed... but I think the biggest shift of all will tilt the scales back in the direction of the PC over the next 12 months when devs and some bleeding edge consumers get their hands on oculus rift dk2. The gaming community has always pushed where the direction of desktop CPU's goes and while it has been going quickly to the tablet of late the desktop is going to be the platform of choice for VR.
VR will save the desktop.

It's not on the way out. The truth is it's bigger than ever.

People have learned the secret, the forbidden fruit, the solution to getting ripped off. People know that assembling their own desktop is as simple as making a pizza!

Pick a good base (motherboard+PSU), add the best tomato sauce (CPU) and cheese (RAM), then throw in plenty of MSG-loaded toppings like pepperoni (GPU), doner meat (SSD) and some chilli for extra tang! :p

Seriously, people build their own these days. Even the biggest morons who later need help getting their drivers and OS sorted out.

I know; I've only ever built my own PCs.

But the fact is, for many people a tablet or smartphone is enough. I don't have numbers for the PC components market, but surely it's very small? Corporates, for one, prefer to buy whole PCs because it comes with support and reduces the chance of driver issues.

That's not true. I've been building my own computers since the 90s and I don't enjoy it. It's easy to run into incompatibilities, for example, between the memory and the motherboard, and when there are so many components that can fail, you'll have a hard time determining which one is causing trouble.

It's not shrinking. The market is still growing; it's just now growing as fast as it used to.

If the demographic grows faster than the market, the market is shrinking.

"the desktop PC is undoubtedly on its way out"

I don't think so. It's hard to assemble a corporate Excel report on an iPhone. Seriously, go to any big firm and announce the end of the Desktop PC. True, there is a definite decline, but I don't see workstations (whether laptop or desktop), going away anytime soon. I think Ubuntu still has a market here - especially if Microsoft doesn't get it's act together.

Why is Conical so deadset on Mir when Wayland is making such good progress?

Speaking of which, is there an equivalent interface for audio? Just wondering.

Why use HTML when we have XML? Sometimes, having flexibility and extensibility is great. Other times, you want something that is specifically designed for a particular task. But Wayland is not a display server. It's a protocol. They'd still have to implement it somehow. But most people shouldn't really care about this. It's a deep technical issue that very few people really understands, though many fight over it. :)

We have two examples of similar things with audio. We have the JACK system and PulseAudio. They're incompatible, but that's hardly ever a problem because they're used in different scenarios.

Using HTML vs XML to analogize Mir vs. Wayland isn't a good comparison. This is partly because HTML and XML are used for completely different things (one is used, generally, for data storage and formatting, the other for structuring and rendering website content).

Canonical made Mir purely out of the desire to have as much control over their software as possible, while at the same time discouraging others from using alternatives, such as Wayland for desktop compositors.

The general understanding among many speculators is that Mir was created solely for business reasons, and any technical justifications made by Canonical was solely for the purpose of maintaining good P.R. with its user base.

http://en.wikipedia.org/wik...

You have no reference for any of your claims. People are just making stuff up and then others believe the conspiracy theory and start spending large resouces in convincing others that the conspiracy must be true.

The idea that the use of GPLv3 is "suspicious" from a company that has always supported the GNU licenses, is just weird.

Just as Ubuntu builds it's desktop and server offerings on top of Debian they could build a tablet/mobile offering on top of Android. You could think of AOSP and Debian as similar levels of abstraction. Where Samsungs TouchWiz is an AOSP based distribution as well as HTC's Sense and others. Why not enter that field? Since both AOSP and Debian are based on the Linux kernel they can also offer a bridge between the two so the two systems can be operating on the same device at the same time and what the user sees would depend on what I/O they were using. Using a touchscreen on a phone or tablet and you get the Android UI. Use a keyboard and mouse on a larger screen and you get a Debian style UI. 64bit multi-core ARM processors, multi-GB or RAM, Full OpenGL/DX gpus on cell phones are on their way so it makes sense that one base OS runs on everything and just displays different UIs. Apple, Google, and Microsoft are likely working toward the same thing.

That's what they did in the early stages. But Android isn't really all that suitable for desktops and they want to have one system for all devices. The solution you're describing would be slow and heavy on resources. You know, this already exists. The reason it's not used, is because it's a suboptimal solution.

You wouldn't be using Android for desktops. As I said in desktop situations you'd be using the Debian/Ubuntu UI. You'd be using the same core linux kernel for both though so you didn't have to have two separate kernels and there would be an overlap in the file system as well (user files accessible in the same place no matter if using the Mobile (android) or Desktop (Ubuntu) based UI). Ubuntu isn't going to unseat Android on mobile. They will never get an app ecosystem that even competes with Amazon's App Store let alone Google Play. But if you can play Google Play apps on your Ubuntu Phone and then plug it to a keyboard, mouse, and monitor, and using the same kernel seamlessly display the Ubuntu Desktop on the monitor (the phone can even still be showing the Android UI) then they'll have a real winner. The only thing I could really see as a threat to that is if Google decides to make an Android based desktop but I've seen no evidence of that. As you said Android isn't really all that suitable for desktop... but Ubuntu is. A hybrid system is a win/win and more elegant then a dual boot solution since both OSs share the same kernel.

That requires a couple of gigabytes of extra ram. Enormous drain on battery and slows things down a great deal. That's why we don't use it. You can do that right now. You have been able to do so for a very long time.

The kernel is important, but it's only a microscopic part of it. You have a window system, a compositor, a shell, etc. There's lots of stuff in a desktop that doesn't exist in phones. Right now, it takes about 40-45 seconds to load on my phone. The idea is that it should be instantaneous, because we're not running a dual system, but one system that adjusts.

Gigabytes of extra RAM? Android required 32MB of RAM until 4.4 where that was raised to only 512MB. Even Android 4.4 devices can run with as little as 340MB but they have to report themselves as "low RAM" devices. Most new phones have 2GB of RAM so unless Ubuntu needs more than 1.5GB there should be plenty of RAM and that's on devices RIGHT NOW. Ubuntu on phones is still years from the mainstream, I believe 16.04 is the current target? With as fast as phone hardware is advancing there should be plenty of gigabytes of RAM available by then. As for battery drain if it's designed intelligently to shut down functions when not in use it shouldn't drain tons of battery.

Sure the kernel isn't everything but using the same kernel without having to unload one to load another is still a big advantage over the dual boot others seem to be experimenting with. Having one system that instantly converts to a mobile platform with no major existing app ecosystem isn't going to appeal to the masses. If Ubuntu can't run Android apps it's not going to go anywhere on mobile beyond the tech-geek circles. Canonical isn't going to be able to compete with Google, Apple, Amazon, and Microsoft. What they can do is leverage their common kernel with Google to carve out a piece of the Android market and since Android is like 80% of mobile their piece of Android may well be more than Microsoft for example has been able to build.

98% of the time I use my laptop as a desktop. If I hand a tablet just as powerful, it would have acted as a desktop as well, desktop usage is not going anywhere.

Congrats, looks really beautiful.

Where is the hardware stats for the build. ? This looks really good. I gave ubantu a swing month ago. Had to search for enabling the sound card which used a 'Sigmatel via command line.

Pcs out the door alright.

Ubuntu is trying to unify different platforms (desktop, laptops, tablets, phones, TVs, ...) in an unique GUI/System. Similar to what Microsoft wanted to do withe Metro UI, but they screwed it because Metro is too much Tablet-ish to be adequate for a desktop.

You have to look at the whole picture, the previews we have seen for tablets, phones and TVs look awesome, they can have something really unique in a year from now.

Distrowatch stats are irrelevant. There are approx 30 million Ubuntu users on this planet. Therefore the few thousand clicks which are counted on the Distrowatch website aren't saying anything about the adoption of a Linux distro. These stats just saying how often a distro page got clicked. That doesn't correlate in any way with being representative for the market share (a few thousand Linux systems would be a poor result). Less known distros or top-ranking distros are also getting more clicks (out of curiosity). Check e.g. out where RedHat is positioned in the ranking and consider which relevance RedHat has for the Linux world. So don't be fooled by the Distrowatch page counter stats.

Ivor, you don't get it. Distrowatch is a great website - content-wise. But a webpage counter of a website is an insufficient statistical method to measure the relevance of a Linux distribution. That understands every child.

I can't find the source where I read that some month ago but some other sites (here one from 2011) may provide some information:
http://www.extremetech.com/...
The Ubuntu website says that there are 20 million users using Ubuntu everyday.
In 2011 was reported that Ubuntu had 12 million users and Fedora having 24 million. The Ubuntu forum has approx. 2 million users. The problem is that Linux downloads are not centralized. Therefore it is difficult to count the amount of Linux distro users. Here the outlook of Canonical: http://www.techspot.com/new... - I assume that will be hard to reach.

This article mentions that 22 million servers get their security update from Canonical (in 2013): http://www.wired.co.uk/news...

Yes - you made your point. In this context I have to confess that I can't take a Linux distribution for serious which is not maintaining its own repositories.

Yea... at least 42 of them clicks is me... I keep loosing my cd every time I install a new copy for someone :P

Tried the latest Linux Mint Cinnamon just last month. Synaptics touchpad on my 2-year old Core i5 Lenovo laptop wasn't even working. Isn't that supposed to be BASIC? So, no go for Mint for me.

But yeah, Ubuntu left a lot of its fans out in the cold when they switched to Unity, me included. And I'm no console command expert.

Hmm, can't remember if I did that or not. Hopefully that's all I need to do to turn the touchpad back on. Oh well. I already chucked the DVD I burned LM on. Thanks anyway, dude.

Mint is collected as one system whereas Ubuntu is split in Ubuntu, Lubuntu, Kubuntu, Xubuntu, Ubuntu Gnome, etc. Count and you'll see that more people are interested in Ubuntu. But that doesn't tell you anything about popularity, because people don't go to read about a system they use every day. They go there to read about the things they _don't_ use.